The Interface of Semantic Interpretation and Inflectional Realization

Abstract

Compositionality is ordinarily conceived of as a syntagmatic notion: a complex expression is semantically compositional if its content can be computed from that of its parts and the manner of their combination. Syntactic structures are held to be largely compositional in this sense. In the same way, a word form may be seen as semantically compositional if its content can be computed from that of its parts and the manner of their morphological combination. I examine this notion in §1. I show that despite the fact that many words can be plausibly seen as semantically compositional, there are many words that plainly are not; I suggest, however, that even these words conform to a distinct, paradigmatic conception of compositionality (§2). This novel view of compositionality suggests that in paradigm-based theories of inflection, a paradigm’s cells have two distinct functions: they serve as a basis for both semantic interpretation and inflectional realization (§3). This assumption is the basis for the cell interface model of the nexus of inflectional morphology and semantics. I show, however, that the cell interface model is overly restrictive—that in cases such as that of Latin deponent verbs, the morphosyntactic property set that determines a paradigm cell’s interpretation is distinct from the set that determines its inflectional realization (§4). Such evidence instead favors the paradigm linkage model, in which semantic composition and inflectional realization proceed from cells of different types; in this model, a central task of a language’s inflectional morphology is that of relating cells of these two sorts (§5). Elaborating on this approach, I analyze a more complex mismatch between content and inflectional form in the inflection of Kashmiri verbs (§6). I briefly examine some wider applications of the paradigm linkage hypothesis (§7), summarizing my conclusions in §8.